A few thoughts on innovation

I was invited the other day to a conference about innovation in the information technology sector. There was nothing remarkable about that event, except perhaps that it led me to voice an opinion I held for years: I do not understand what people are really talking about when they talk about innovation, at least in software, that is.

It might be odd to write this, but if there’s any concept that’s both fuzzy and dangerously misleading in the software industry, that would be innovation. I have read for many years and listened to people explaining how to “stirr and create innovation” in a company or in a community. Maybe these words have been used for lack of a better term; but I still don’t see how you can create innovation. I think you might be able to stirr it somehow, as it’s already a humbler verb. But frankly, can someone out there tell me what does innovation mean in the software world?

In general terms, I would define innovation as the big and small changes constantly leading to a change of the art in any given field. I think that’s pretty much what one usually understands by that word. So why could this not be applied to software? Precisely because software is rarely -if at all- the result of big changes happening all of a sudden and by accident. Software development usually happens at an incremental pace, whether openly so (think about the agile development practices) or even when there’s a structured corporate environment favoring traditional code reviews and quality assurance processes through stable product development cycles. Software is not produced by accident. Software is the result of process, and in theory accidents do not happen there. In fact, I could also point out that incremental changes or a period of technological incubation might be observed right before the emergence of almost any given technology. Take the medieval rudder for instance: it’s been rumored to have been imported in Europe around the 12th century by Chinese ships, but there are tracks and evidence of previous try-outs by European sailors and shipyards to design wooden rudders and articulate them with a complete mechanism. Similarly, it is hard to say how “innovation” happened in the sixties when the U.S. decided to send manned flights to the moon, but the wave of small and not so small innovation that was the result of this huge project is still visible to everyone (think of the Tefal pans, among many other things).

Thus there are, I think, two points that need to be highlighted: First, innovation does not happen all of a sudden if the field of software field and more generally ICT. It is a set of processes that ultimately lead to new software, or software that’s supposedly not as bad as the former state of the art. Second, what’s unclear is how -to quote several people I listened to- innovation “happens”. It sounds sometimes that innovation is a mystery or the philosophers’ stone that require care and secrecy to happen. Yet in the software industry, it does not work that way, for all the marketing and bells and whistles that come out of software vendors do not brush aside the fact that even inside these corporations software development is a set of very well defined, but non-public, processes.

Innovation is not a mystery and I don’t think that you can track how it works. You can assume that a certain set of circumstances and an environment letting people code start-ups emerge and Free & Open Source Software projects grow will ultimately translate into something that someone, whether a journalist, consultant, politicians or venture capitalists will call innovation. Anything else besides that, innovation sounds more like vapor and magical boxes. This should probably express what I feel about software patents, by the way.

One last thing: Innovation is different than progress. Progress is usually applied to fields that do not necessarily belong to science or technology; it can be more a perception and may concern society as a whole. Yet the interesting thing is that while progress seems to be an even more elusive term than innovation, you can actually tell progress from regression or stagnation: people perceive it almost immediately, however relative it sometimes may be.

Enjoy the beginning of the Holiday season!